cgroup-utils 0.5 (beta)

Utility tools for control groups of Linux

For users

$ sudo pip install cgroup-utils

For developers

$ git clone git://github.com/peo3/cgroup-utils.git
$ cd cgroup-utils
$ python setup.py build
$ sudo python setup.py install

$ python setup.py bdist --formats=rpm

cgutil configs

This command show you configurations of cgroups. By default, it shows only changed configurations.

$ cgutil configs -o memory
<root>
    notify_on_release=1
    release_agent=/usr/lib/ulatencyd/ulatencyd_cleanup.lua
sys_essential
    swappiness=0
    notify_on_release=1
sys_bg
    swappiness=100
    notify_on_release=1

cgutil event

This command makes cgroup.event_control easy to use. It exits when a state of a target cgroup crosses a threshold which you set, thus, you can know the state of the cgroup has changed.

$ cgutil event /sys/fs/cgroup/memory/system/sshd.service/memory.usage_in_bytes +1M
$ # It exits when memory usage of processes in the cgroup has increased one more MB.

cgutil pgrep

This command is alike pgrep command but it shows cgroups in addtion to PIDs.

$ cgutil pgrep ssh
/: 15072
/: 15074
/system/sshd.service: 630
$ cgutil pgrep ssh -l -f
/: 15072 sshd: ozaki-r [priv]
/: 15074 sshd: ozaki-r@pts/2
/: 15157 /bin/python /bin/cgutil pgrep ssh -l -f
/system/sshd.service: 630 /usr/sbin/sshd -D

cgutil stats

This command shows you states of cgroups.

$ cgutil stats
<root>
        stat={'throttled_time': 0L, 'nr_periods': 0L, 'nr_throttled': 0L}
system
        stat={'throttled_time': 0L, 'nr_periods': 0L, 'nr_throttled': 0L}
system/sm-client.service
        stat={'throttled_time': 0L, 'nr_periods': 0L, 'nr_throttled': 0L}
system/sendmail.service
        stat={'throttled_time': 0L, 'nr_periods': 0L, 'nr_throttled': 0L}
system/vboxadd-service.service
        stat={'throttled_time': 0L, 'nr_periods': 0L, 'nr_throttled': 0L}
system/colord.service
        stat={'throttled_time': 0L, 'nr_periods': 0L, 'nr_throttled': 0L}
system/colord-sane.service
        stat={'throttled_time': 0L, 'nr_periods': 0L, 'nr_throttled': 0L}
system/udisks2.service
        stat={'throttled_time': 0L, 'nr_periods': 0L, 'nr_throttled': 0L}
system/cups.service
        stat={'throttled_time': 0L, 'nr_periods': 0L, 'nr_throttled': 0L}

cgutil top

This command is alike top command but it shows activities in a unit of cgroups.

$ cgutil top -i -n 2 -b
18.1 msec to collect statistics
[  CPUACCT  ]  [     BLKIO     ]  [        MEMORY       ]
 USR    SYS      READ    WRITE     TOTAL    RSS     SWAP     # NAME
  0.0%   0.0%    0.0 /s   0.0 /s     0.0    48.0k    0.0    97 usr_1000/default
20.5 msec to collect statistics
[  CPUACCT  ]  [     BLKIO     ]  [        MEMORY       ]
 USR    SYS      READ    WRITE     TOTAL    RSS     SWAP     # NAME
  0.0%   0.0%    0.0 /s   0.0 /s   128.0k    4.0k    0.0   104 sys_daemon
  0.0%   0.0%    0.0 /s   0.0 /s   -64.0k    0.0     0.0     0 sys_essential
  0.0%   0.0%    0.0 /s   0.0 /s   108.0k   32.0k    0.0    97 usr_1000/default

cgutil tree

This command shows you tree structure of cgroups.

$ cgutil tree -o blkio
<root>
   `system
       +sm-client.service
       +sendmail.service
       +vboxadd-service.service
       +colord.service
       +colord-sane.service
       +udisks2.service
       +cups.service
       +rtkit-daemon.service

(snip)

       +fsck@.service
       +udev.service
       `systemd-journald.service

Supported subsystems

• blkio (and its debug feature)
• cpuset
• cpu and cpuacct
• devices
• freezer
• hugetlb
• memory
• net_cls
• net_prio

0.5

• Support hugetlb subsystem
• Support additional cgroup files of blkio, memory and freezer
• Use argparse instead of optparse
• Fix several bugs. Thanks novas0x2a and unicell!

0.4

• Add unit tests for control file parsers and a test for PEP8 compliance
• Support dynamic cgroups hierarchy update for top command
• Support missing net_prio subsystem
• Complete event_control supports in event command
• Implement mkdir/rmdir commands
• Fix a bunch of bugs